A New Study on Medically Related Bankruptcies

Thanks to our friends over at WSJ's Bankruptcy Beat, a new study caught my eye on the issue of medical bankruptcies. A new study appearing in the Journal of Clinical Oncology documents an increased risk of bankruptcy with certain types of cancers. The full abstract is available.

The study is principally directed at understanding what contributes to bankruptcy risk as between different types of cancers. But, if we can use cancer as an indicator of serious medical problems, the numbers can be used to draw some comparisons between medical problems and general bankruptcy risk. The conclusions provide some support for both sides of the debate about whether medical problems lead to an increased risk of bankruptcy.

Conditioned on survival to the point in time, the study finds lung cancer patients have a 1.5% probability of filing bankruptcy within one year of their diagnosis. For thyroid cancer or lymphoma/leukemia patients, the one-year probabilities are 0.9% and 0.6%. At five years, the bankruptcy probabilities are 7.7% for lung cancer, 4.8% for thyroid cancer, and 3.6% for lymphoma/leukemia. Thus, although it is impossible to say just from the abstract, the trend over time appears to be close to linear. Stated differently, the probability of filing bankruptcy increases about an equal amount each year such that the probability of filing at any time in the first two years is about twice as much as the first year alone, that the probability of filing at any point in the first three years is about three times as much the first year alone, and so on.

The cancer study used data from Washington state. Right now, according to data from Epiq Systems, the bankruptcy filing rate in the state of Washington is 4.96 bankruptcy petitions per thousand persons. That is not too far off the national annual filing rate of 4.85 bankruptcy petitions per thousand persons. Of the bankruptcy petitions in Washington, a high end estimate from national data might be that 10% are short-term repeat filers and (using more reliable data than is available on repeat filers) about 30% are joint petitions involving a husband and wife. To make apples-to-applies comparisons between the bankruptcy data and the data from the cancer we should subtract the short-term repeat filers to avoid double-counting and add the joint petitions. If we do so, a rough estimate is that the probability of anyone in the state of Washington filing bankruptcy in a given year is a little over 0.6%.

Comparing the incidence of bankruptcy in the general population to the incidence of bankruptcy among cancer patients does not lead to a clear conclusion. Some cancers (lung and thyroid) lead to a greater risk of bankruptcy than the general population while others (uterine, colorectal, and melanoma) appear to have about the same or slightly lower risks of bankruptcy. Prostate cancer patients appear to have a lower risk of bankruptcy. The dispersion could be random -- some cancer types had to be above or below the mean -- but the increased rate of bankruptcy among lung cancer patients is fairly large making for a low probability of random variation in the data. Are there spending patterns associated with these cancers that account for differences in bankruptcy? Are lung cancer patients, for example, more likely to have unreimbursed medical expenses? One controversial hypothesis that the data can't prove would be that lung cancer and bankruptcy are both associated with risky behavior that reflect some underlying personality type.

Another finding from the study is that cancer patients over the age of 65 had a lower probability of bankruptcy than younger cancer patients. According to the abstract, the authors speculate that the reason might be the availability of Medicare, leading to fewer unreimbursed medical debts among the elderly. Medicare is likely playing a role, but I also would point out that the elderly are more likely to be using Social Security and pension benefits as sources of income. These income sources have greater protections from creditor garnishment (although sometimes more in theory than practice), and are thus perhaps less likely to necessitate a bankruptcy filing for protection of income. Also, although among demographic groups seniors have seen the highest rate of increase in bankruptcy filings, their bankruptcy rate still remains low as compared to other age groups.

Having just read back over what I wrote, the only message seems to be that the conclusions one draws from this study need to be nuanced and careful. I really need to work on this blogging thing.